Biomarkers are cellular, biochemical or molecular alterations that can be easily and non-invasively measured in human tissues and are directly or indirectly in the pathway of disease. Over the last decade, scientific research has repeatedly shown that ambient air particulate pollution and lead exposures accelerate cognitive aging. The limited availability of biomarkers that reflect at-risk exposures and cognitive decline dramaticall limits opportunities for effective targeted prevention. To address this gap, our long-term goal is to identify novel mitochondriomic biomarkers that can reconstruct past environmental influences and predict the risk of future cognitive impairment. We will use state-of-art blood-based analyses of 1) mitochondrial DNA (mtDNA) damage and aging by means of a panel of complementary markers of oxidative damage, heteroplasmy and mtDNA abundance; and 2) methylation of nuclear DNA (nDNA) genes coding for mitochondrial proteins (for which we will have the unique opportunity to time- and cost-effectively abstract information from ready-to-use genome- wide DNA methylation data). These markers have properties that make them exceptionally well suited to biomarker development, as they: i) have been shown to be persistently altered in blood by environmental exposures; and ii) can mark the presence of aged, damaged blood mitochondria, a primary source of systemic oxidative stress to which the brain is particularly vulnerable. A major challenge for age-related biomarkers is that - precisely because they change over time - cross-sectional and case-control studies, which offer more potential to analyze relevant target/diseased tissue samples, cannot provide evidence of the temporality of associations. Altered biomarkers could merely reflect lifestyle changes or other influences that followed, rather than preceded cognitive deficits. In this proposal, we exploit the unique nature of the Normative Aging Study (NAS) cohort in which we have repeated collections of DNA samples, exposure and cognitive data every 3.5 years for 2 decades. Because of this, we can determine whether biomarker changes, measured longitudinally over two time points, both follow exposure and precede disease. This is a major methodological advantage over an approach that utilizes a case-control design. All findings will be independently validated in MOBILIZE, a cohort remarkably similar to the NAS for study design, exposure levels, and participants' characteristics. We will confirm DNA methylation findings by Pyrosequencing, and we will analyze mRNA expression to assess the function of the DNA methylation changes. We will use advanced statistical modeling to integrate a panel of biomarkers of systemic oxidative stress and inflammation in the paths linking exposure, blood mitochondriomics, and cognitive decline. As we can leverage a wealth of extant resources and data from the NAS and MOBILIZE cohorts, we will be able to cost-effectively recapitulate data from up to two decades of aging over the cycle of a 4-year grant.